CHAPTER 3 RESEARCH DESIGN AND METHODS
The propose of this chapter is to provide a description of the research design and the methods that are used to attain the objectives of the study and answer the research questions. The researcher also describes the research design and methods, sampling techniques, data collection procedures, data analysis, and validity and reliability of the research data.
The research design is a plan used as the basis for decision-making that spells out the basic strategies to develop information that is accurate and interpretable. It is a logical model that links research to inquiry and methods for collecting and analysing empirical data (Thupayagale-Tshweneagae 2008:88). The research design also guides the investigators on how to collect, analyse and interpret observations. The main purpose of selecting an appropriate study design is to reduce possible errors and bias by maximising the reliability and validity of data (Mukhopadhyay 2007:283-284; Christian, Daniel & Jens 2009:3-8).
There are several study designs in research. For this study, a pre-test-post-test design with non- equivalent control group, known as quasi-experimental design was used. In this design, the groups of study participants are naturally assembled into two groups, one being the experimental group that receives treatment, and the other the control group.
For this study, both groups took the same pre-test in order to establish initial equivalence, thus obtain the baseline data and identify gaps associated with HIV and AIDS-related SAD. After indentifying the gaps, the treatment group took a short course training related to HIV, AIDS, and SAD whereas the control group did not take the training. The control group helps to determine whether or not the intervention actually made a difference in the treatment group. Finally, the two groups took the same post-test four weeks after the training. Then after, the relative effectiveness of the intervention was assessed by comparing the observed difference between the post-tests of the treatment and control groups.
The advantage of this type of design is that it controls for history, maturation, testing, and instrumentation as sources of invalidity. The design has also the advantage of convenience, feasibility and practicability. Its disadvantage is that since the units are not randomly assigned to treatment and control groups, selection bias can never be excluded completely resulting in reduced internal validity. In order to substantially reduce the bias, the characteristics of the respondents in both groups were determined before the intervention (Jonassen 2008:1023-1026; Reichardt 2009; Chris 2012:79-83.).
APPROACHES USED IN THE STUDY
There two approaches used in this study are quantitative and qualitative.
The quantitative approach uses various strategies of inquiry and methods for data collection and analysis (Thupayagale-Tshweneagae 2008:89). The emphasis of quantitative paradigm is on collecting and analysing numerical data using descriptive and/or inferential statistics. This type of paradigm, although harder to design in the beginning, is usually very much detailed and structured, and its findings can be easily collated and presented statistically (Colin 2007:3; Robin 2009:4-8).
This paradigm is more subjective in nature than the quantitative one and it mainly deals with exploring and reflecting intangible naturalistic phenomena such as values, beliefs, perceptions and attitudes. When used in conjunction with the quantitative paradigm, it can help to interpret and understand the complex reality of a given situation. In spite of the fact that this paradigm is easier to start, it is more difficult to interpret and present the findings (Colin 2007:3; Thupayagale-Tshweneagae 2008:88; Robin 2009:4-8).
Polit and Beck (2010:246) found that qualitative research is considered pragmatic, interpretive but it should be sensitive to the social context in which the data are collected. Symond and Gorard (2010:4) pointed out that the two approaches complement each other when merged, the quantitative one being objective compared to the subjectivity of the qualitative approach.
The present study has employed both quantitative and qualitative approaches. The use of multiple sources enables the triangulation and validation of information for robust and reliable conclusions.
The study was conducted in two referral hospitals found in two cities in Amhara Region. The Amhara region is one of the regions in Ethiopia, located in the northwest part of the country, with currently estimated population of 20,398,999. The capital city of the region is Bahir Dar, 563 kilo meters far away from Addis Ababa, The capital city of Ethiopia, just near to the Lake Tana. In the region, there are 19 public hospitals out of which five are referral hospitals. The referral hospitals provide outpatient, in-patient and emergency services including all HIV and AIDS-related services, like HCT and comprehensive HIV and AIDS treatment, care and support.
Population and sample selection
Target and study population
The target population for the proposed study were all HCPs working in the referral hospitals and the study population were HCPs working in the two randomly selected referral hospitals. The accessible population were physicians, health officers, midwives, nurses and laboratory professionals selected from the two hospitals. For the qualitative part of the study, PLWHA admitted as in-patients during the study period were the study responednts.
Participants and respondents included in the study had to meet this criteria:
Permanently employed physicians, health officers, midwifes, nurses and laboratory technicians.
HCPs who had served less than six months in the referral hospitals.
Sampling method and technique for the quantitative approach
The sampling method for the quantitative approach was probability sampling. In this type of sampling, a particular sample from a specified population has a known (non-zero) probability of being selected. This method of sampling is more likely to result in a representative sample. In this sampling method, each element of the study population has equal chance of being included in the sample. The sampling technique chosen for this study was stratified random sampling in which the sampling fraction was the same for each stratum. In this technique, the study population is subdivided into subpopulation called strata having similarity within them, but with distinctive differences between the strata. Then, the respondents were selected using stratified random sampling technique from each stratum. Finally, estimates of each stratum are combined to produce an estimate for total sample size (Patrick 2008:3-8; Mukhopadhyay 2007:98-113; Bowling 2009:203-212).
The advantage of stratified random sampling technique is that it is more representative and the sample can be kept in small size without losing its accuracy. In addition, the characteristics of each stratum can be easily estimated and compared. The drawbacks of this technique are that it is costly to prepare stratified lists of all members and there may be problems of faulty classification that will result in increase in variability (Mukhopadhyay 2007:102-103).
The main reason for selecting the two referral hospitals was that the number of HCPs and the HIV case load is relatively higher than that of the general and district hospitals.
This has enabled the investigator to obtain adequate sample size and valid information. From the selected hospitals, HCPs mostly dealing with PLWHA by virtue of their profession were divided into strata and then using stratified random sampling technique, the study participants were selected from each stratum. Accordingly, 110 HCPs were randomly selected with 95% confidence level and 5% margin of error (Mukhopadhyay 2007:106-113; Patrick 2008:3-68; Caldwell 2011:1-4).
Sampling method and technique of the qualitative approach
The sampling method for this study was non-probability sampling. In this method, sample elements are chosen by non-random technique as proposed by Mukhopadhyay (2007:98-113); Bowling (2009:203-212) and Rao and Richard (2006:58-65).
Purposive sampling technique was used to collect qualitative data. The advantages of purposive sampling are that it is simple to draw, less costly and involves less field workers. Its disadvantages are that it does not yield representative sample as it is not randomly selected. Moreover, it requires a considerable knowledge about the study population (Bowling 2009:206-208).
Quantitative data collection
Purposive sampling was also used for the quantitative aspect of the study. The data collection method was structured self-administered questionnaires. The structured questionnaires were adapted from different literature sources and contextualised to Ethiopian health care settings. The HCPs from the two hospitals were given the same questionnaire both in the pre-test and post-test phases. After getting their informed consent, the study participants were allowed to complete the questionnaire in private settings. The questionnaires were designed to collect data, including knowledge about HIV, AIDS, HIV and AIDS-related to SAD, perceived risks, attitudes, beliefs and practices towards PLWHA. The questionnaire also included hospital policies, guidelines and protocols related to HIV and AIDS-related SAD in the health care settings.
Procedure of the quantitative data collection
After getting permissions from concerned authorities, two data collectors were assigned to each hospital to facilitate the data collection processes both for the pre-tests and post-tests. Two supervisors were also assigned to monitor and oversee all data collection processes. Each day, the collected data were checked by the supervisors and the investigator for consistency and completeness. The data collectors and the supervisors were recruited based on their educational background and previous experiences. The data collectors were those HCPs with Bsc. Degree in health sciences while the supervisors were health professionals with Master Degree in public health. In order to ensure, the data collectors and the supervisors have attended two-day training sessions regarding on how to scrutinise and manage the data collection processes.
Pre-testing of the questionnaire
Polit and Beck (2009:336) indicated that pre-testing of the questionnaire is used to identify and detect any gaps inherent in the instruments and to determine the effectiveness of the tools prior to the data collection.
For this study, the structured questionnaire was pre-tested on 20 HCPs having similar characteristics with the intended respondents. The respondents of the pre-testing were from other referral hospital which was not part of the main study. The pre-testing was used to determine whether or not the questions are relevant to study objectives and the respondents have clearly understood the questions. Moreover, the pre-testing was aimed at ensuring whether or not the wordings and sequences of the questions are correct and consistent. The pre-testing of the tool also enabled the investigator to ensure the completeness and consistencies of the research instruments. The main flaws identified during the pre-testing were inappropriate wording and sequence of the questionnaire. Based up on the gaps indentified through the pre-testing, the questionnaires were revised and prepared for the actual data collection (Mukhopadhyay 2007:287).
Qualitative data collection
In-depth interview (IDI) was used to obtain the qualitative data. The IDI is one of the qualitative data collection methods that involve conducting individual interview with a small number of clients to explore their beliefs, feelings, attitudes and practices on a particular programme and situations. Its main advantage is that it provides much more detailed information. Information through the use of this method can also be used to augment the validity and reliability of the quantitative data. Some of the limitations the IDI are that it is prone to bias, time intensive and not generalisable to target population (Mukhopadhyay 2007:287-289; Jonassen 2008:1059; USAID 2008b:11; Hancock, Ockleford & Windridge 2009:16-17; Bowling 2009:407-415).
The qualitative data were obtained through in-depth interview with HIV-positive patients admitted to each referral hospital. The IDI was carried out using guiding questions translated into local language (Amharic). Accordingly, PLWHA who were admitted to the two selected hospitals were interviewed in private settings until the data saturation point occurs (Hancock et al 2009:22).
To conduct the IDI at each hospital, two qualified and experienced health care professionals were recruited and assigned after getting one-day training on how to conduct the IDIs. Patients, whose condition is not improved, were excluded from the IDI. The responses of the patients were noted and audio-taped after obtaining informed consents from each interviewee. After the IDI, the data collectors thanked the respondents for taking their time.
Statistical analysis enables the investigator to reduce, compile, organise, analyse, evaluate and interpret the data. The primary purpose of data analysis is to provide appropriate answers to research questions that will enable the investigator to attain the intended objectives of the study. Prior to the data analysis, there is a need to process the collected data using appropriate statistical package (Wu et al 2008).
The data processing should preferably begin as early as possible. It mainly includes the data entry, cleaning and editing. The processing of the data is intended to check for consistency, errors and incompleteness. Prior to the data processing, the data should be scrutinised properly. The data processing is a crucial stage and if it is well-planned, it can result in a rapid statistical analysis. It is also critical to keep track of the data collected in a condition which would facilitate a quick review of the progress and quality of the data management (Mukhopadhyay 2007:5).
For this study, the data processing started at the time of the data collection. Prior to the data collection, the questionnaire was categorised and coded in order to enhance the process of the data collection and analysis. Accordingly, a template for data entry was carefully developed in consultation with the data analyst. After the data entry, data cleaning and editing was carried out by the investigator.
Quantitative data analysis
For this study, the data analysis was carried out by using Statistical Packages for Social Sciences (SPSS) version 20.0 and Microsoft Excel to generate tables and figures. The statistical data analysis involves descriptive and inferential statistics.
Descriptive data analysis
Descriptive data analysis enables the researcher to synthesise and describe quantitative data obtained from empirical observations and measurements. This analysis includes univariate and bi-variate analyses, The univariate analysis is intended to describe data for one variable while the bi-variate data analysis is concerned with describing the magnitude and existence of relationships between two variables using tabular and diagrammatic presentations (Mukhopadhyay 2007:2-54; Dakhale, Hiware, Shinde & Mahatme 2012:435-442).
For this study, descriptive data analysis of the pre-tests of both treatment and control groups was executed for socio-demographic variables, knowledge about HIV and AIDS, attitudes and beliefs of HCPs towards PLWHA, fears and perceived risk of infection, health facility environment and practices as well as hospital polices, guidelines and protocols.
Inferential data analysis
Inferential statistics is used to draw conclusions about a characteristics of a population based upon the data obtained from the sample. With inferential statistics, the investigator estimates the population parameters from the sample statistics. A statistical inference deals with estimation of parameters. The estimation procedures are used to estimate a single population parameter. The chi-square (X2) is used for categorical variables to test difference in proportions in two or more groups. There are also advanced type of inferential procedures that include multivariate analysis, linear, multiple and logistic regressions. Multivariate analysis is a statistical procedure that enables the investigator to understand the effects of two or more independent variables on a dependent variable that is continuous in nature. Regression analysis is mostly performed to make predictions about phenomena. In linear regression, one independent variable is used to predict a dependent variable. Multiple regression is used when there are more than two or more independent variables. Logistic regression is procedure that is employed to analyse the relationships between multiple independent and dependent variables that are categorical in nature (Mukhopadhyay 2007:116-145; Christian, Daniel & Jens 2009:1-12; Dakhale et al 2012:435-442).
CHAPTER 1 ORIENTAION OF THE STUDY
1.2 BACKGROUND TO THE RESEARCH PROBLEM
1.3 STATEMENT OF THE PROBLEM
1.4 THE THEORETICAL FRAMEWORK
1.5 DEFINITONS OF KEY CONCEPTS
1.6 AIM OBJECTIVES AND RESEARCH
1.8 SIGNIFICANCE OF THE STUDY
1.9 ETHICAL CONSIDERATIONS
1.10 SCOPE THE STUDY
1.11 STRUCTURE OF THE THESIS
CHAPTER 2 LITERATURE REVIEW
2.2 THEORETICAL FRAMEWORK
2.3 HIV AND AIDS
2.4 STIGMA AND DISCRIMANTION
2.5 FACTORS ASSOCIATED WITH HIV AND AIDS-RELATED SAD IN HEALTH CARE SETTINGS
2.6 EXPERIENCES OF HIV AND AIDS-RELATED SAD IN THE HEALTH CARE SETTINGS
2.7 IMPACT OF HIV AND AIDS-RELATED SAD IN THE HEALTH CARE SETTINGS
2.8 HIV AND AIDS-RELATED SAD-REDUCTION INTERVENTIONS
2.9 EFFECTIVE STRATEGIES TO REDUCE HIV AND AIDS-RELATED STIGMA AND DISCRIMINATION
2.10 THE ETHIOPIAN HEALTH CARE SYSTEM
CHAPTER 3 RESEARCH DESIGN AND METHODS
3.2 RESEARCH DESIGN
3.3 APPROACHES USED IN THE STUDY
3.4 RESEARCH METHODS
3.5 BRIEF INTERVENTION
3.6 DATA QUALITY
3.7 ETHICAL CONSIDERATIONS
3.8 DISSEMINATION OF THE RESULTS
CHAPTER 4 FINDINGS OF THE STUDY
4.2 FINDINGS OF THE QUALITATIVE PARADIGM
CHAPTER 5 BRIEF INTREVENTION
CHAPTER 6 DISCUSSION
6.2 MAGNITUDE OF THE HIV AND AIDS-RELATED SAD IN THE HOSPITALS
6.3 EFFECTS OF THE INTERVENTION ON THE RESPONDENTS
6.4 DESIGNING HIV AND AIDS-RELATED SAD REDUCTION-INTERVENTION STRATEGY
CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS
7.2 SUMMARY AND INTERPRETATION OF THE STUDY FINDINGS
7.5 CONTRIBUTION OF THE STUDY
7.6 LIMITATIONS OF THE STUDY
LIST OF REFERENCES
GET THE COMPLETE PROJECT
HIV AND AIDS-RELATED STIGMA AND DISCRIMINATION REDUCTION-INTERVENTION STRATEGY IN HEALTH CARE SETTINGS OF AMAHARA REGION, ETHIOPIA